Squaraine dyes containing diphenylamine group: Effects of different type structures on material properties and organic photovoltaic performances

Article abstract of DOI:10.1016/j.dyepig.2019.107763

Four donor-acceptor-donor (D-A-D’) unsymmetrical squaraines (USQs) with different molecular skeletons (XZ-type and YZ-type), containing diphenylamine group with/without methoxy substituent were synthesized as donor materials in bulk-heterojunction (BHJ) organic photovoltaics (OPVs). The introduction of methoxy group in USQs has little different effects on the overall photovoltaic performance. Conversely, the different molecular skeleton types of the USQs have significant influence on their material properties and photovoltaic performances. Compared to BIDPSQ and BIDPOMeSQ with XZ-type molecular skeleton, IDPSQ and IDPOMeSQ with YZ-type molecular skeleton display closer solid-state packing, much lower highest occupied molecular orbital (HOMO) energy level, higher hole mobility and smaller phase separation domain size. Consequently, YZ-type USQs exhibit the most excellent performance with power conversion efficiency (PCE) of ~4%, which is approximately 300% higher than those of XZ-type USQs. Surprisingly, even though IDPSQ and IDPOMeSQ show wide band gaps, the corresponding devices still achieve a highest PCE of ~4%, which is comparable to or even higher than the PCEs of some reported SQ-based devices with low band gaps. These results indicate the YZ-type molecular skeleton and the electron-donating diphenylamine group are very promising to construct highly efficient squaraine donor materials.

Full text of DOI:10.1016/j.dyepig.2019.107763

Journal Pre-proof
Squaraine dyes containing diphenylamine group: Effects of different type structures
on material properties and organic photovoltaic performances
Lin Yang, Youqin Zhu, Jianglin Wu, Bin Hu, Zhenguo Pang, Zhiyun Lu, Suling Zhao,
Yan Huang
PII:
S0143-7208(19)31369-5
DOI:
https://doi.org/10.1016/j.dyepig.2019.107763
DYPI 107763
Reference:
To appear in: Dyes and Pigments
Received Date: 12 June 2019
Revised Date: 27 July 2019
Accepted Date: 30 July 2019
Please cite this article as: Yang L, Zhu Y, Wu J, Hu B, Pang Z, Lu Z, Zhao S, Huang Y, Squaraine
dyes containing diphenylamine group: Effects of different type structures on material properties
and organic photovoltaic performances, Dyes and Pigments (2019), doi: https://doi.org/10.1016/
j.dyepig.2019.107763.
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition
of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of
record. This version will undergo additional copyediting, typesetting and review before it is published
in its final form, but we are providing this version to give early visibility of the article. Please note that,
during the production process, errors may be discovered which could affect the content, and all legal
disclaimers that apply to the journal pertain.
© 2019 Published by Elsevier Ltd.

Table of Contents artwork:
The YZ-type USQs exhibit the power conversion efficiency of ~ 4%, which is
approximately 300% higher than those of XZ-type USQs.

Squaraine dyes containing diphenylamine group: Effects of different type
structures on material properties and organic photovoltaic performances
Lin Yang, a, b, 1 Youqin Zhu, c, 1 Jianglin Wu,a Bin Hu,a Zhenguo Pang,a Zhiyun Lu,a Suling Zhao, c, ** Yan
Huang a, *
a
Key Laboratory of Green Chemistry and Technology (Ministry of Education), College of Chemistry,
Sichuan University, Chengdu 610064, P. R. China. E-mail: huangyan@scu.edu.cn
b School of Pharmaceutical Sciences, Southwest Medical University, Luzhou, Sichuan 646000, P. R. China
c
Key Laboratory of Luminescence and Optical Information (Ministry of Education), Institute of
Optoelectronics Technology, Beijing Jiaotong University, Beijing 100044, P. R. China. E-mail:
slzhao@bjtu.edu.cn
1 These authors contributed equally.

Abstract:
Four donor-acceptor-donor (D-A-D’) unsymmetrical squaraines (USQs) with different molecular skeletons
(XZ-type and YZ-type), containing diphenylamine group with/without methoxy substituent were
synthesized as donor materials in bulk-heterojunction (BHJ) organic photovoltaics (OPVs). The introduction
of methoxy group in USQs has little different effects on the overall photovoltaic performance. Conversely,
the different molecular skeleton types of the USQs have significant influence on their material properties
and photovoltaic performance. Compared to BIDPSQ and BIDPOMeSQ with XZ-type molecular skeleton,
IDPSQ and IDPOMeSQ with YZ-type molecular skeleton display closer solid-state packing, much lower
highest occupied molecular orbital (HOMO) energy level, higher hole mobility and smaller phase separation
domain size. Consequently, YZ-type USQs exhibit the most excellent performance with power conversion
efficiencies (PCE) of ~ 4%, which is approximately 300% higher than those of XZ-type USQs. Surprisingly,
even though IDPSQ and IDPOMeSQ show wide band gaps, the corresponding devices still achieve a highest
PCE of ~ 4%, which is comparable to or even higher than the PCEs of some reported SQ-based devices with
low band gaps. These results indicate the YZ-type molecular skeleton and the electron-donating
diphenylamine group are very promising to construct highly efficient squaraine donor materials.
Keywords: squaraines; diphenylamine; organic photovoltaics; molecular skeletons; structure and
performance

1. Introduction
Organic photovoltaics (OPVs) have attracted considerable interest as a kind of promising technology for
renewable solar energy applications due to their cost-effectiveness and large-area flexible fabrication.[1-7]
Recent advances have pushed the power conversion efficiencies (PCE) of OPVs to over 15%,[5, 8-10]
which is inseparable from the material innovation, device optimization and fundamental understanding of
device physics. Among them, the development and design of donor-acceptor (D-A) conjugated donor and
acceptor materials plays a particularly important role in progress toward increased PCEs of OPVs.[4, 11-13]
The core strategy of developing D-A conjugated photovoltaic materials is to select appropriate D or A units,
and then combine D and A units in different ways, such as D-A, D-A-D, A-D-A, D-A-A, D-D-A,
D-A-D-A-D or A-D-A-D-A, etc. and then modify side chains to balance the absorption, energy level, carrier
mobility and phase morphology of the photoactive layer. At present, most of the highest-performance donor
and acceptor materials are generally constructed with conjugated backbones that have alternating D and A
aromatic units, such as PBDB-TF, ITIC etc.[8, 14, 15]
Squaraines (SQs), a class of dyes with resonance stabilized zwitterionic structure, are synthesized from
the condensation of squaric acid with two electron-rich moieties to form a donor-acceptor-donor (D-A-D)
skeleton.[16] Recently, SQs have emerged as promising donor materials in OPVs due to their characteristics
of strong absorption in the visible and near-infrared regions, simple synthetic routes, remarkable stability
and wide molecular structure diversity.[17-22] According to the different kinds of electron-rich moieties,
SQs can pertain to three general structures illustrated in Scheme 1.[18, 19, 23-26] X-type squaraines are the
condensation products of squaric acid with appropriate anhydrobases,[24, 27] while Y-type squaraines are
prepared starting from activated arenes or π-excessive heterocycles,[18, 28-31] Z-type squaraines are
prepared starting from primary or secondary amines.[32, 33] The three types of SQs show very divergent
photophysical, chemical, electric and molecular packing properties.[25, 34-36] Among the SQ photovoltaic
materials, the Y-type SQs are the main ones, and show the obviously higher PCEs (4 ~ 6%) than those of

X-type SQs (1 ~ 2%).[19] Relatively speaking, the Z-type SQs are rather rare, and only two cases of
unsymmetrical YZ-type SQs (ASSQ and DPASQ, shown in Scheme 2), which contain diphenylamine group
as electron-donating moiety, were reported to be employed as donor materials in planar heterojunction (PHJ)
OPVs with the power conversion efficiency (PCE) of ~ 4%.[18, 35]
Scheme 1. Three types of photovoltaic squaraine dyes (D = electron-rich aromatic or heterocyclic rings; R =
alkyl or aryl groups).[23]
Scheme 2. The molecular structures of ASSQ and DPASQ.
In fact, due to the good electron-donating and hole transporting capabilities of diphenylamine (DPA)
group, many π-conjugated oligomers and small molecules containing diphenylamine in the bridging or
terminal positions have been widely employed as the active ingredient in optical and electrical materials.[37,
38] Therefore, very recently, we also synthesized a novel unsymmetrical YZ-type SQ (IDPSQ, shown in
Scheme 3) containing diphenylamine group, which was employed as donor materials in solution-processing
bulk heterojunction (BHJ) OPVs with PCE of ~ 4%.[23] Furthermore, IDPSQ were also successfully
applied to ternary solar cells and semitransparent solar cells, and the corresponding photovoltaic devices
achieved excellent performance.[39, 40]
In this work, based on the molecular structure of IDPSQ (YZ-type), the other three unsymmetrical
squaraines (USQs, namely BIDPSQ, BIDPOMeSQ and IDPOMeSQ, shown in Scheme 3) with XZ-type or
YZ-type skeleton, containing diphenylamine or p-methoxy substituted diphenylamine as electron-donating

moiety, were synthesized to research the effects of the molecular structures of these different type SQs on
their molecular properties and photovoltaic performances. The results show that the different type structures
have effectively influence on their solid-state packing, energy level, carrier mobility and photovoltaic
performance. Consequently, YZ-type SQs (IDPSQ and IDPOMeSQ) exhibit a highest PCE of ~ 4%, which
is approximately 300% higher than those of XZ-type SQs (BIDPSQ and BIDPOMeSQ). In addition, the
presence or absence of methoxy groups also has some slight effect on molecular properties and device
performance.
Scheme 3. Molecular structures of the target USQs and IDPSQ.
2. Experimental section
2.1. Measurements
1H and 13C NMR spectra were recorded on a Bruker Avance AV ΙΙ 400 MHz instrument with
tetramethylsilane as internal standard. High resolution mass spectra (HRMS) were measured on a Shimadzu
LCMS-IT-TOF. Absorption spectra of USQs in 5 × 10-6 mol L-1 chloroform solution and thin film states
were measured with a PerkinElmer Lambda 950 UV-Vis scanning spectrophotometer. The thin film samples
were fabricated by spin-casting the USQ chloroform solution with a concentration of 5 mg mL-1 on quartz
substrates (1500 rpm, 30 s).
Cyclic voltammetry was performed in 5 × 10-4 mol L-1 USQ anhydrous dichloromethane solution with a
LK 2010 electrochemical work station, using a three-electrode cell with a Pt disk working electrode, a Pt
wire counter electrode and a Ag/AgNO3 (0.1 mol L-1 in acetonitrile) reference electrode.

Tetrabutylammonium perchlorate (0.10 mol L-1) and ferrocenium/ferrocene redox couple were used as the
supporting electrolyte and internal potential reference, respectively.
The crystallographic data for IDPSQ is obtained from the literature (CCDC 1487656)[23]. The
crystallographic data for BIDPSQ and BIDPOMeSQ have been deposited in the Cambridge Database
(CCDC 1487658 for BIDPSQ, and 1903751 for BIDPOMeSQ). Single crystal samples of BIDPSQ and
BIDPOMeSQ were obtained from dichloromethane/methanol system. Single crystal X-ray diffraction data
of BIDPSQ were obtained on a Xcalibur E X-ray single crystal diffractometer equipped with graphite
monochromator Mo-Kα (λ = 0.7107 Å) radiation. The data collection was executed using CrysAlisPro
program. The structures were determined using direct method and successive Fourier difference syntheses
(SHELXS-97) and refined using full-matrix least-squares procedure on F2 with anisotropic thermal
parameters for all non-hydrogen atoms (SHELXL-97). Single crystal X-ray diffraction data of BIDPOMeSQ
were obtained on a Gemini X-ray single crystal diffractometer equipped with graphite monochromator
Cu-Kα (λ = 1.5418 Å) radiation. The refinement details and the resulting factors for these USQs are given in
Table S1.
Samples for atomic force microscopy (AFM) measurements were prepared by spin-casting from
USQ:PC71BM = 1:3 in chloroform solution with a total concentration of 20 mg mL-1 on glass substrates
(1500 rpm, 50 s). The transmission electron microscopy (TEM) investigation was performed on a FEI
Tecnai
G2
F20
S-TWIN
field
emission
transmission
electron
microscope.
The
poly(3,4-ethylenedioxythiophene)-poly(styrenesulfonate) (PEDOT:PSS) layer was prepared by spin-casting
o
their water solution on the glass substrate (3000 rpm, 40 s) and then baked at 150 C for 10 min. The
specimen for TEM measurement was prepared by spin-casting the blend chloroform solution (USQ:PC71BM
= 1:3, 20 mg mL-1) on the glass/PEDOT:PSS substrate, then floating the film on water surface, and
transferring to copper grids.
2.2. Device preparation

Photovoltaic devices were fabricated with a layered structure of glass/ ITO/ MoO3 (8 nm)/ USQ:PC71BM/
LiF (0.8 nm)/ Al (100 nm). The ITO-coated glass substrates (sheet resistance = 15 Ω sq-1) were cleaned
through sequential sonication in detergent, deionized water, acetone and ethanol for 10 min each, and finally
blow-dried by high purity nitrogen. The substrates were treated by UV-ozone for 5 min, then immediately
transferred into a high vacuum chamber for deposition of 8 nm MoO3 at pressure of < 3 × 10-4 Pa with a rate
of 0.5 Å s-1. Subsequently, the photoactive layers were fabricated by spin-casting the USQ:PC71BM
chloroform solution with a total concentration of 20 mg mL-1 (1500 rpm, 50 s) in a N2-filling glove box at
25 oC. Then the substrates were loaded into a vacuum chamber to finish the deposition of LiF (0.8 nm) and
Al (100 nm) under pressure of < 3 × 10-4 Pa with a rate of 0.05 Å s-1 and 1.5 Å s-1, respectively. Deposition
rate and film thickness were in situ monitored using a quartz crystal oscillator mounted to the substrate
holder. The active area of cells was 6 mm2. To obtain the average data related to device performance, several
batches of devices (4 cells per batch) for each set of conditions were fabricated and tested. The current
density-voltage curves under illumination were measured using an Abet solar simulator with a Keithley 4200
source measurement unit under AM 1.5G illumination (100 mW cm-2), after spectral mismatch correction
o
under an ambient atmosphere at 25 C. EQE measurements were performed in air using a QE/IPCE
Measurements Solar Cell Scan 100 (ZOLIX) system.
Hole-only devices were fabricated with structure of ITO/ MoO3 (8 nm)/ USQ (80 nm) or USQ:PC71BM
(1:3, wt%, 80 nm)/Au (100 nm). The dark current density-voltage characteristics of these devices were
measured and fitted the results using the space charge limited current (SCLC) model. The current density (J)
is given by
ꢈ
ꢀ = ꢂꢁ ꢃꢄꢃꢅꢆ ꢇꢉ
ꢊ
Where ε0 is the permittivity of free-space, εr is the relative dielectric constant of the active layer, µ is the
charge carrier mobility, and L is the thickness of the active layers.
2.3. Synthesis

The synthetic routes of intermediates and USQs are outlined in Scheme 4. Compounds IDPSQ and 1-8 were
prepared according to the procedures described in the literatures.[18, 23, 41-43] The other chemicals,
reagents, and solvents were used as received from the suppliers except as specifically mentioned.
Scheme 4. Synthetic routes of three USQs. Reagents/conditions: a) isopropanol, concentrated H2SO4, reflux,
3 h, 81%; b) diphenylamine, isopropanol, concentrated HCl, reflux, 8 h, 48%; c) acetone, 6 mol L-1 HCl,
reflux, 8 h, 55% (3), 73% (6); d) NaOBu-t (1.5 eq.), Pd(OAc)2 (4% eq.), P(t-Bu)3HBF4 (10% eq.),
anhydrous toluene, reflux under Ar, 5 h, 72%; e) 1, isopropanol, concentrated HCl, reflux, 8 h, 48%; f)
quinoline, n-butanol/benzene = 4:1 (v/v), reflux, 13 h, 37% (BIDPSQ); g) quinoline, n-butanol/toluene = 1:3
(v/v), reflux, 13 h, 44% (BIDPOMeSQ); h) n-butanol/toluene = 1:3 (v/v), reflux, 13 h, 45% (IDPOMeSQ).
2.3.1. 3,4-Diisopropoxycyclobut-3-ene-1,2-dione (1)[18]
A mixture of squaric acid (2.00 g, 17.5 mmol), isopropanol (20 mL) and a drop of concentrated H2SO4
was refluxed for 3 h, and the solvent was evaporated under reduced pressure. Then isopropanol was added,
the mixture was refluxed for 30 min and the solvent was evaporated again. The crude product was purified

by silica gel column chromatography (petroleum ether:ethyl acetate = 8:1) to afford 1 (2.80 g, 81%) as white
solid. M.p. 42-45 oC. 1H NMR (400 MHz, CDCl3, ppm) δ 5.39-5.29 (m, 2H, CH), 1.45 (d, J = 6.4 Hz, 12H,
CH3).
2.3.2. 3-(Diphenylamino)-4-isopropoxycyclobut-3-ene-1,2-dione (2)[18]
A mixture of diphenylamine (5.64 g, 33.3 mmol), 1 (6.00 g, 30.3 mmol), isopropanol (200 mL) and
concentrated HCl (0.6 mL) was refluxed for 8 h, and the reaction was monitored by thin layer
chromatography. Then reaction mixture was concentrated in vacuo. The resulting crude product was purified
by silica gel column chromatography (petroleum ether:ethyl acetate = 10:1) to afford 2 (4.54 g, 48%) as
yellow solid. M.p. 156-158 oC. 1H NMR (400 MHz, CDCl3, ppm) δ 7.38 (t, J = 7.2 Hz, 4H, ArH), 7.30 (t, J
= 7.2 Hz, 2H, ArH), 7.11 (d, J = 7.6 Hz, 4H, ArH), 5.46-5.37 (m, 1H, CH), 1.28 (d, J = 6.0 Hz, 6H, CH3).
2.3.3. 3-(Diphenylamino)-4-hydroxycyclobut-3-ene-1,2-dione (3)[18]
2 (4.54 g, 14.7 mmol) was added to acetone (100 mL), stirred and dissolved. After adding 6 mol L-1 HCl
(150 mL), the reaction mixture was refluxed for 8 h, and the reaction was monitored by thin layer
chromatography. The reaction mixture was poured into water, and the yellow precipitate was obtained by
filtration. The precipitate was recrystallized from acetone to afford 3 (2.15 g, 55%) as pale-yellow solid. M.p.
o
1
186-187 C. H NMR (400 MHz, DMSO-d6, ppm) δ 7.29 (t, J = 7.2 Hz, 4H, ArH), 7.10 (t, J = 7.2 Hz, 2H,
ArH), 6.95 (d, J = 8.0 Hz, 4H, ArH).
2.3.4. Bis(4-methoxyphenyl)amine (4)[41]
4-Methoxybromobenzene (2.00 g, 10.7 mmol), 4-methoxyaniline (1.32 g, 10.7 mmol), Pd(OAc)2 (96 mg,
4% eq.), P(t-Bu)3.HBF4 (310 mg, 10% eq.), sodium tert-butoxide (3.08 g, 32.1 mmol) were added to
anhydrous toluene (50 mL), and the mixture was heated at 110 °C for 5 h, then filtered. The filtrate was
evaporated under reduced pressure, and the crude product was purified by silica gel column chromatography
1
(petroleum ether:ethyl acetate = 10:1) to afford 4 (2.04 g, 72%) as white solid. M.p. 102-103ºC. H NMR

(400 MHz, CDCl3, ppm) δ 6.92 (br, 1H, NH), 6.83 (d, J = 7.2 Hz, 8H, ArH), 3.78 (s, 6H, CH3).
2.3.5. 3-(Bis(4-methoxyphenyl)amino)-4-isopropoxycyclobut-3-ene-1,2-dione (5)[28]
4 (0.80 g, 3.49 mmol), 1 (0.63 g, 3.17 mmol) were added to the mixture of isopropanol (30 mL) and
concentrated HCl (0.1 mL). The mixture was refluxed for several hours, and the reaction was monitored by
thin layer chromatography. When the reaction mixture was cooled, the solvent was evaporated under
reduced pressure. The crude product was purified by silica gel column chromatography (petroleum
1
ether:ethyl acetate = 10:1) to afford 5 (0.65 g, 59%) as grey solid. M.p. 162-163ºC. H NMR (400 MHz,
CDCl3, ppm) δ 7.03 (dt, J = 9.2 Hz, 3.2 Hz, 2.4 Hz, 4H, ArH), 6.87 (dt, J = 9.2 Hz, 3.2 Hz, 2.4 Hz, 4H,
ArH), 5.46-5.37 (m, 1H, CH), 3.82 (s, 6H, CH3), 1.30 (d, J = 6.0 Hz, 6H, CH3).
2.3.6. 3-(Bis(4-methoxyphenyl)amino)-4-hydroxycyclobut-3-ene-1,2-dione (6)[37]
The preparation method is the same as 3. After reaction, the mixture was poured into water to precipitate a
viscous solid, and the supernatant was decanted. The viscous solid was dissolved in dichloromethane, and
petroleum ether was added dropwise to precipitate a gray precipitate, filtered to afford 6 (0.44 g, 73%) as
pale-yellow solid. M.p. 216-218ºC. 1H NMR (400 MHz, CDCl3, ppm) δ 7.09 (dt, J = 9.2 Hz, 3.2 Hz, 2.4 Hz,
4H, ArH), 6.89 (dt, J = 9.2 Hz, 3.2 Hz, 2.4 Hz, 4H, ArH), 3.83 (s, 6H, CH3).
2.3.7. (E)-2-((3-butyl-1,1-dimethyl-1,3-dihydro-2H-benzo[e]indol-2-ylidene)methyl)-4-(diphenyliminio)-3-
oxocyclobut-1-en-1-olate (BIDPSQ)
3 (1.00 g, 3.8 mmol), 7 (1.63 g, 3.8 mmol), mixed solvent (benzene: n-butanol = 1:4, v/v, 150 mL) and
quinoline (0.6 mL) were refluxed for 13 h. Then the solvent was evaporated under reduced pressure and the
crude product was purified by column chromatography (dichloromethane/methanol) and followed by
recrystallization from dichloromethane/methanol to afford BIDPSQ as fuchsia crystal (0.85g, 37%). M.p.
251-252 ºC. 1H NMR (400 MHz, CDCl3, ppm) δ 8.12 (d, J = 8.4 Hz, 1H, ArH, H9), 7.89 (d, J = 8.8 Hz, 1H,
ArH, H6), 7.87 (d, J = 9.2 Hz, 1H, ArH, H5), 7.55 (t, J = 7.6 Hz, 1H, ArH, H7), 7.45-7.40 (m, 5H, ArH, H2,

H8), 7.35-7.29 (m, 3H, ArH, H3, H4), 7.26 (t, J = 3.6 Hz, 4H, ArH, H1), 6.07 (s, 1H, CH, H10), 4.12 (t, J =
6.4 Hz, 2H, CH2, H12), 2.00 (s, 6H, CH3, H11), 1.87-1.79 (m, 2H, CH2, H13), 1.53-1.42 (m, 2H, CH2, H14),
0.99 (t, J = 7.6 Hz, 3H, CH3, H15 ). 13C NMR (100 MHz, CDCl3, ppm) δ 184.3, 174.1, 173.2, 140.8, 139.4,
134.5, 131.4, 129.8, 129.0, 128.7, 127.5, 127.4, 125.5, 124.6, 122.6, 110.3, 86.1, 51.5, 43.8, 29.5, 27.0, 20.4,
+
14.0. HRMS (ESI)+ m/z: [M+H]+ calcd. C35H33N2O2 , 513.2537; found, 513.2532.
2.3.8.
(E)-4-(bis(4-methoxyphenyl)iminio)-2-((3-butyl-1,1-dimethyl-1,3-dihydro-2H-benzo[e]indol-2-
ylidene)methyl)-3-oxocyclobut-1-en-1-olate (BIDPOMeSQ)
6 (300 mg, 0.92 mmol), 7 (363 mg, 0.92 mmol), mixed solvent (toluene: n-butanol = 3:1, v/v, 20 mL) and
quinoline (0.2 mL) were refluxed for 13 h. Then the solvent was evaporated under reduced pressure and the
crude product was purified by column chromatography (dichloromethane/methanol) and followed by
recrystallization from dichloromethane/methanol to afford BIDPOMeSQ as fuchsia crystal (0.25g, 44%).
M.p. 276-278 ºC. 1H NMR (400 MHz, CDCl3, ppm) δ 8.11 (d, J = 8.4 Hz, 1H, ArH, H9), 7.88 (d, J = 8.4 Hz,
1H, ArH, H6), 7.85 (d, J = 8.8 Hz, 1H, ArH, H5), 7.53 (td, J = 7.6 Hz, 1.2 Hz, 1H, ArH, H7), 7.39 (td, J = 7.6
Hz, 1.2 Hz, 1H, ArH, H8), 7.27 (d, J = 8.0 Hz, 1H, ArH, H4), 7.18 (dt, J = 8.8 Hz, 3.6 Hz, 2.0 Hz, 4H, ArH,
H1), 6.93 (dt, J = 8.8 Hz, 3.6 Hz, 2.0 Hz, 4H, ArH, H2), 5.96 (s, 1H, CH, H10), 4.06 (br, 2H, CH2, H12), 3.84
(s, 6H, CH3, H3), 1.99 (s, 6H, CH3, H11), 1.85-1.78 (m, 2H, CH2, H13), 1.51-1.42 (m, 2H, CH2, H14), 0.99 (t,
J = 7.2 Hz, 3H, CH3, H15).13C NMR (100 MHz, CDCl3, ppm) δ 182.1, 174.2, 172.1, 158.8, 139.6, 134.2,
131.3, 129.8, 129.7, 128.7, 127.4, 126.5, 124.3, 122.5, 114.2, 110.2, 85.8, 55.6, 51.2, 43.6, 29.4, 27.0, 20.4,
+
14.0. HRMS (ESI)+ m/z: [M+H]+ calcd. C37H37N2O4 , 573.2748; found, 573.2741.
2.3.9. 4-(Bis(4-methoxyphenyl)iminio)-2-(2,6-dihydroxy-4-(indolin-1-yl)phenyl)-3-oxocyclobut-1-en-1-olate
(IDPOMeSQ)
6 (440 mg, 1.35 mmol), 8 (300 mg, 1.32 mmol) and the mixed solvent (toluene: n-butanol = 3:1, v/v, 20
mL) were refluxed for 13 h. Then the solvent was evaporated under reduced pressure and the crude product

was purified by column chromatography (dichloromethane/petroleum ether) and followed by
recrystallization from dichloromethane/methanol to afford IDPOMeSQ as brown solid (0.27g, 45%). M.p.
249-250 ºC. 1H NMR (400 MHz, CDCl3, ppm) δ 11.89 (s, 2H, OH, H4), 7.41 (d, J = 8.0 Hz, 1H, ArH, H6),
7.21 (d, J = 7.2 Hz, 1H, ArH, H9), 7.19-7.14 (m, 5H, ArH, H1, H7), 6.96 (dt, J = 8.8 Hz, 3.6 Hz, 2.0 Hz, 4H,
ArH, H2), 6.92 (td, J = 7.2 Hz, 0.8 Hz, 1H, ArH, H8), 6.26 (s, 2H, ArH, H5), 4.03 (t, J = 8.0 Hz, 2H, CH2,
H11), 3.86 (s, 6H, CH3, H3), 3.14 (t, J = 8.0 Hz, 2H, CH2, H10).13C NMR (100 MHz, CDCl3, ppm) δ 176.5,
175.9, 166.8, 163.4, 159.8, 153.7, 143.8, 133.4, 132.9, 127.5, 126.6, 125.5, 122.3, 114.5, 113.5, 105.0, 96.5,
+
55.7, 52.1, 28.1. HRMS (ESI)+ m/z: [M+H]+ calcd. C32H27N2O6 , 535.1864; found, 535.1867.
3. Results and discussion
3.1. Synthesis and characterization
The synthetic routes of BIDPSQ, BIDPOMeSQ and IDPOMeSQ are outlined in Scheme 4. Compound
IDPSQ and intermediates 1-8 were synthesized according to the literature. [18, 23, 41-43] The molecular
1
structures of these USQs were confirmed by H NMR, 13C NMR, HR-ESIMS and X-ray single
crystallographic analysis. Moreover, all of the USQs show good solubility in chloroform at room
temperature (> 15 mg mL-1), which is essential to preparing smooth, and uniform films through solution
processing.
3.2. Single crystal X-ray crystallography
The crystallographic data for IDPSQ is obtained from the literature (CCDC 1487656).[23] The single
crystals of BIDPSQ and BIDPOMeSQ were obtained by solvent evaporation from
dichloromethane/methanol system. The corresponding crystal data have been deposited in the Cambridge
Database (CCDC 1487658 for BIDPSQ, and 1903751 for BIDPOMeSQ). Unfortunately, the single crystal
of IDPOMeSQ wasn’t obtained from various solvent systems such as dichloromethane/methanol,
dichloromethane/n-hexane,
dichloromethane/cyclohexane,
chloroform/n-hexane
and

chloroform/cyclohexane. Actually, when purifying the four USQs, the BIDPSQ, BIDPOMeSQ and IDPSQ
were crystals with metallic luster by recrystallization from dichloromethane/methanol system, while
IDPOMeSQ was brown powdery solid with no metallic luster. Therefore, the optimal conditions and method
for cultivating the single crystal of IDPOMeSQ may be quite different from the other three USQs, which
needs more time to explore. The relevant crystal data and crystal structure refinement parameters are shown
in Table S1.
The single crystals of BIDPSQ, BIDPOMeSQ and IDPSQ are monoclinic, and the space groups are P21/c,
P21/c, and P21/n, respectively. The molecular ellipsoid (ORTEP), corresponding dihedral angle and
inter-molecular packing are shown in Fig. 1. As expected, the two phenyl rings of the diphenylamine group
in the three USQs have a large dihedral angle which is greater than 70°. Comparing BIDPSQ and
BIDPOMeSQ, it can be found that the bond lengths on the diphenylamine group of BIDPOMeSQ are
slightly shortened, and the corresponding dihedral angle is slightly reduced. BIDPOMeSQ shows a more
planar structure than that of BIDPSQ, which might be attributed to the electron-donating ability of methoxy
substituent. Therefore, the shortest interplane π…π distance of BIDPOMeSQ is 4.206 Å, smaller than that of
BIDPSQ (4.528 Å), indicating that the introduction of methoxy substituent is beneficial to promote effective
intermolecular packing. On the other hand, comparing BIDPSQ and IDPSQ, it can be found that the
squarate ring and the benzindole unit in BIDPSQ are almost coplanar (the dihedral angle is only 3.30o). In
IDPSQ, the torsional angle between squarate ring and dihydroxyphenyl group is 7.34o, additionally, the
dihedral angle between the dihydroxyphenyl group and the indoline segment is as large as 27.16o. However,
the shortest π…π distance in IDPSQ single crystal is 3.864 Å, which is much shorter than that of BIDPSQ
(4.528 Å). Moreover, the density of IDPSQ single crystal is calculated to be 1.407 mg mm-3, larger than that
of BIDPSQ single crystal (1.214 mg mm-3). These results demonstrate that the intermolecular stacking of
IDPSQ (YZ-type SQ) is confirmed to be more effective than that of BIDPSQ (XZ-type SQ), which should
be propitious to the enhancement of charge carrier mobility. In addition, in IDPSQ, the bond lengths of

O-H…O are calculated to be 1.765 and 1.822 Å (Fig. 1), indicative of the presence of relatively strong
intramolecular hydrogen bonding interactions between the two hydroxyl groups of dihydroxyphenyl group
and oxygen atoms on the central squarate ring.[44] According to our previous research result that the
introduction of intramolecular hydrogen bond interactions in SQs is benefit to enhance the compatibility of
SQ donor materials and [6,6]-phenyl-C71butyric acid methyl ester (PC71BM) then decrease the phase
separation domain size of their blend film,[45] it could be deduced that IDPSQ-based blend film might
obtain a smaller phase separation size than that of BIDPSQ-based one.
Fig. 1. The molecular ellipsoid (ORTEP), corresponding dihedral angle and inter-molecular packing models
of BIDPSQ, BIDPOMeSQ and IDPSQ.
3.3. Optical properties
The UV-visible absorption spectra of BIDPSQ, BIDPOMeSQ, IDPSQ and IDPOMeSQ in dilute chloroform
solution (5.0 × 10-6 mol L-1) and thin films are shown in Fig. 2. The corresponding data are summarized in
Table 1. In the dilute chloroform solution, all USQs have similar strong absorption in the visible region, and

the maximum absorption wavelength (λmax) is located at 548 ~ 564 nm, with a relatively high molar
extinction coefficient (ε, 7.36 ~ 9.90 × 104 L mol-1 cm-1). Comparing the absorption spectra of two groups of
USQs with or without methoxy substituent (BIDPSQ vs. BIDPOMeSQ, IDPSQ vs. IDPOMeSQ), it can be
seen that the introduction of methoxy substituent causes a slight red shift in the absorption spectrum of the
corresponding USQ, which may be attributed to the electron-donating ability of methoxy substituent,
consistent with the above conclusion from the single crystal analysis. Meanwhile, the absorption spectra of
IDPSQ and IDPOMeSQ have a red shift of 10 ~ 12 nm when compared with BIDPSQ and BIDPOMeSQ. In
the film state, all USQs show the broadening and red shift absorption spectra with λmax located at 572 ~ 582
nm. According to the onset of the absorption spectra in solid-state films, the optical band gap (ꢋꢌꢍꢎꢏ) of
BIDPSQ, BIDPOMeSQ, IDPSQ and IDPOMeSQ are estimated to be 1.95 eV, 1.93 eV, 1.83 eV and 1.86 eV,
respectively.[46] It means these USQs belong to the wide band gap donor materials, and has potential to
prepare promising ternary organic photovoltaics with other low band gap materials.
a)
b)
Film
1.0
0.8
0.6
0.4
0.2
0.0
1.0
0.8
0.6
0.4
0.2
0.0
Solution
BIDPSQ
BIDPSQ
BIDPOMeSQ
IDPSQ
BIDPOMeSQ
IDPSQ
IDPOMeSQ
IDPOMeSQ
300
400
500
Wavelength (nm)
600
700
400
500
600
Wavelength (nm)
700
800
Fig. 2. UV-visible absorption spectra of BIDPSQ, BIDPOMeSQ, IDPSQ and IDPOMeSQ in dilute
chloroform solution (a) and solid films (b).
3.4. Electrochemical properties
The highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) energy
levels of USQs were determined via cyclic voltammetry (CV) measurements (shown in Fig. 3 and Table 1).
In the anode scan, all these USQs show two-electron oxidation processes in the range of 0.2 ~ 1 V, and the

onset of the first oxidation wave vs. Fc/Fc+ is located at 0.29 V for BIDPSQ, 0.21 V for BIDPOMeSQ, 0.56
V for IDPSQ, and 0.52 V for IDPOMeSQ. According to the energy level of Fc/Fc+ redox couple (4.80 eV
below vacuum),[46] the HOMO energy levels of BIDPSQ, BIDPOMeSQ, IDPSQ and IDPOMeSQ are
calculated to be -5.09 eV, -5.01 eV, -5.36 eV and -5.32 eV, respectively. In the cathode scan, these USQs
show an irreversible reduction oxidation process in the range of -1 ~ -1.6 V, and the onset of the reduction
wave vs. Fc/Fc+ is located at -1.68 V for BIDPSQ, -1.67 V for BIDPOMeSQ, -1.39 V for IDPSQ, and -1.43
V for IDPOMeSQ. The LUMO energy levels of BIDPSQ, BIDPOMeSQ, IDPSQ and IDPOMeSQ are
calculated to be -3.12 eV, -3.13 eV, -3.41 eV, and -3.37 eV, respectively. Obviously, the HOMO and LUMO
levels of BIDPOMeSQ (IDPOMeSQ) with methoxy substituent are higher than those of the corresponding
methoxy-free BIDPSQ (IDPSQ), which is attributed to the electron-donating ability of methoxy substituent.
In addition, unexpectedly, the HOMO and LUMO levels of IDPSQ (IDPOMeSQ) are actually ~ 0.30 eV
lower than the corresponding BIDPSQ (BIDPOMeSQ), which may be due to the formation of
intramolecular hydrogen bonding between the hydroxyl groups and the oxygen atoms on the central squarate
ring in IDPSQ (IDPOMeSQ) reducing the overall energy of the whole π-conjugated system.[45, 47]
Table 1 Optical data and energy levels of the four USQs.
solutiona
filmb
ꢋꢌꢍꢎꢏ
Eox
(V)d
Ered
(V)d
HOMO
(eV)e
LUMO
(eV)e
Compound
λmax (nm,
ε, 104 M-1 cm-1)
λmax
(nm)
(eV)c
1.95
1.93
1.83
1.86
BIDPSQ
BIDPOMeSQ
IDPSQ
548 (7.79)
552 (7.36)
558 (9.56)
564 (9.90)
572
576
578
582
0.29 -1.68
0.21 -1.67
0.56 -1.39
0.52 -1.43
-5.09
-5.01
-5.36
-5.32
-3.12
-3.13
-3.41
-3.37
IDPOMeSQ
a Measured in CHCl3, c = 5.0 × 10-6 mol L-1.
b Thin films spin-casted from 5 mg mL-1 CHCl3 solution on quartz substrates.
c Obtained from the onset of UV-Vis absorption spectra in the thin film state.
d Redox potential values were measured by cyclic voltammetry vs. Fc/Fc+.
e HOMO = (-4.8-qꢋꢍꢐ) eV, LUMO = (-4.8-qꢋꢅꢑꢒ) eV.[46]

-1500-1000 -500
IDPOMeSQ
0
500 1000
IDPSQ
BIDPOMeSQ
BIDPSQ
-1500-1000 -500
0
500 1000
Voltage (mV)
Fig. 3. Cyclic voltammograms of the four USQs in CH2Cl2 solution.
3.5. Photovoltaic properties
To research the photovoltaic performance of these USQ compounds, BHJ-OPV cells using USQ as electron
donor material and PC71BM as electron acceptor material were fabricated by solution processing with a
device structure of ITO/ MoO3 (8 nm)/ USQ:PC71BM (wt/wt, 80 nm)/ LiF (0.7 nm)/ Al (100 nm). Initially,
the effect of different blend ratios of USQ:PC71BM on the photovoltaic performance of the devices were
investigated, and the relative current density-voltage (J-V) curves and photovoltaic data are shown in Table 2
and Fig. 4. For these USQ-based systems, they display different trend in the dependence of photovoltaic
performance on the USQ:PC71BM blend ratio. For BIDPSQ-based device, when the BIDPSQ:PC71BM blend
ratio is varied from 1:1 to 1:3, the PCE is enhanced from 0.29% to 0.90% along with significantly increased
Jsc and FF and slightly reduced Voc. When the blend ratio is further varied from 1:3 to 1:8, the photovoltaic
performances of the devices just show slight increment (0.90% ~ 1.10%) with increased Jsc and almost
unchanged FF and Voc. The optimal BIDPSQ:PC71BM blend ratio is found to be 1:8, and the resultant device
shows the highest PCE of 1.10% with Jsc = 4.50 mA cm-2, Voc = 0.71 V, and FF = 0.35. For
BIDPOMeSQ-based device, when the blend ratio is varied from 1:1 to 1:3, the PCE is enhanced from 0.31%
to 0.71% along with increased Jsc and FF and slightly reduced Voc. When the blend ratio is further varied

1
0
1
0
a)
b)
-1
-2
-3
-4
-5
-6
-1
-2
-3
-4
-5
BIDPSQ:PC71BM
BIDPOMeSQ:PC71BM
1:1
1:3
1:5
1:8
1:1
1:3
1:5
1:8
0.0
0.2
0.4
0.6
0.8
1.0
0.0
0.2
0.4
0.6
0.8
1.0
Voltage (V)
Voltage (V)
c)
d)
0
0
-2
-4
-6
-8
-2
-4
-6
-8
IDPSQ:PC71BM
IDPOMeSQ:PC71BM
1:1
1:1
1:3
1:5
1:8
1:3
1:5
1:8
-10
-10
0.0
0.2
0.4
0.6
0.8
1.0
0.0
0.2
0.4
0.6
0.8
1.0
1.2
Voltage (V)
Voltage (V)
e)
BIDPSQ
USQ:PC71BM=1:3
BIDPOMeSQ
IDPSQ
0.5
IDPOMeSQ
0.4
0.3
0.2
0.1
0.0
300
400
500
600
700
800
Wavelength (nm)
Fig. 4. The J-V curves (a-d) of BIDPSQ-, BIDPOMeSQ-, IDPSQ-, IDPOMeSQ-based devices under
different USQ:PC71BM blend ratio and the EQE curves (e) of the devices with USQ:PC71BM = 1:3.
from 1:3 to 1:8, the photovoltaic performances of the devices are almost unchanged (0.69% ~ 0.72%), since
the Jsc, FF and Voc change one after another. The optimal BIDPOMeSQ:PC71BM blend ratio is also 1:8, and
the resultant device shows the highest PCE of 0.72% with Jsc = 3.27 mA cm-2, Voc = 0.67 V, and FF = 0.31.
For IDPSQ and IDPOMeSQ-based devices, when the blend ratio is varied from 1:1 to 1:8, they show a
similar trend in photovoltaic performance. Taking IDPSQ-based device as an example, when the blend ratio
is varied from 1:1 to 1:3, the Voc of the device decreases, but the corresponding Jsc and FF significantly
increase, resulting in an increase in PCE from 2.76% to 3.20%. When the blend ratio is varied from 1:3 to

1:8, the Voc of the device is continued to decrease, and the corresponding Jsc and FF also decrease
significantly, resulting in a decrease in efficiency from 3.20% to 1.70%. Finally, the optimal blend ratio of
the IDPSQ-based device is 1:3, and the highest PCE is 3.20% with Jsc = 9.02 mA cm-2, Voc = 0.95 V, and FF
= 0.38. It is worth mentioning that the IDPSQ-based device achieves a Voc of up to 1.00 V when the blend
ratio is 1:1, which is related to the rather deep HOMO level (-5.34 eV) of IDPSQ. Similarly, the optimal
blend ratio of the IDPOMeSQ device is also 1:3 with a highest PCE of 3.23% with Jsc = 9.15 mA cm-2, Voc =
0.88 V, and FF = 0.40.
In order to compare the device performance of the four systems, the photovoltaic performances of the
blend ratio of 1:3 were discussed. Apparently, according to the photovoltaic data in Table 2, the
Table 2 The photovoltaic performance data of USQ-OPV devices with varied USQ:PC71BM blend ratio.
Voc
Jsc
FF
PCE
(%)
Donor
(V)
(mA cm-2)
1:1 0.75 (0.72±0.02) 1.39 (1.28±0.12) 0.31 (0.32±0.01) 0.32 (0.29±0.03)
1:3 0.69 (0.70±0.02) 3.76 (3.61±0.15) 0.36 (0.36±0.01) 0.95 (0.90±0.05)
1:5 0.70 (0.69±0.01) 4.10 (4.08±0.10) 0.35 (0.35±0.01) 1.01 (0.99±0.03)
1:8 0.70 (0.71±0.01) 4.87 (4.50±0.37) 0.35 (0.35±0.01) 1.21 (1.10±0.11)
1:3a 0.76 (0.74±0.02) 3.89 (3.78±0.11) 0.36 (0.36±0.01) 1.06 (1.00±0.06)
1:1 0.72 (0.73±0.01) 1.37 (1.24±0.13) 0.34 (0.34±0.01) 0.34 (0.31±0.03)
1:3 0.71 (0.70±0.01) 2.96 (2.81±0.17) 0.37 (0.36±0.01) 0.78 (0.71±0.07)
1:5 0.68 (0.66±0.02) 3.74 (3.30±0.44) 0.36 (0.32±0.03) 0.91 (0.69±0.22)
1:8 0.68 (0.67±0.01) 2.90 (3.27±0.34) 0.41 (0.33±0.06) 0.81 (0.72±0.08)
1:3a 0.68 (0.69±0.01) 3.38 (3.31±0.13) 0.38 (0.37±0.01) 0.87 (0.84±0.03)
1:1 1.01 (1.00±0.01) 7.88 (7.63±0.25) 0.36 (0.36±0.01) 2.90 (2.76±0.14)
1:3 0.99 (0.95±0.03) 9.24 (9.02±0.22) 0.37 (0.38±0.01) 3.42 (3.20±0.22)
1:5 0.88 (0.89±0.01) 8.22 (7.78±0.44) 0.35 (0.34±0.01) 2.56 (2.39±0.17)
1:8 0.94 (0.90±0.03) 6.20 (6.06±0.14) 0.31 (0.31±0.01) 1.82 (1.70±0.12)
1:3a 0.96 (0.96±0.01) 9.20 (9.22±0.21) 0.39 (0.37±0.02) 3.42 (3.30±0.12)
1:1 0.98 (0.97±0.01) 7.33 (7.12±0.21) 0.39 (0.37±0.02) 2.81 (2.57±0.24)
1:3 0.89 (0.88±0.01) 9.41 (9.15±0.26) 0.42 (0.40±0.02) 3.48 (3.23±0.25)
1:5 0.87 (0.84±0.02) 8.78 (8.53±0.25) 0.39 (0.38±0.01) 2.95 (2.78±0.17)
1:8 0.78 (0.78±0.01) 7.66 (7.47±0.19) 0.35 (0.35±0.01) 2.10 (2.02±0.08)
1:3a 0.93 (0.94±0.01) 9.77 (9.63±0.14) 0.45 (0.43±0.02) 4.06 (3.93±0.13)
BIDPSQ
BIDPOMeSQ
IDPSQ
IDPOMeSQ
a Thermally annealed at 90 oC for 10 min.

IDPSQ-based and IDPOMeSQ-based devices perform significantly better than the BIDPSQ-based and
BIDPOMeSQ-based ones, mainly due to the much higher Voc and Jsc of the first two devices. The Voc of
IDPSQ and IDPOMeSQ devices are 0.95 V and 0.88 V, respectively, which are much higher than those of
BIDPSQ and BIDPOMeSQ devices (~ 0.70 V), which is mainly due to the fact that the HOMO levels of the
former two USQs are deeper than the latter two. Meanwhile, the Jsc of the IDPSQ and IDPOMeSQ devices
are 9.02 mA cm-2 and 9.15 mA cm-2, respectively, which is also much higher than those of BIDPSQ and
BIDPOMeSQ devices (3.61 mA cm-2, 2.81 mA cm-2). It is consistent with the corresponding EQE curves
(Fig. 4e). Obviously, the EQE values of IDPSQ and IDPOMeSQ devices are much higher than BIDPSQ and
BIDPOMeSQ devices in the entire region (300 ~ 750 nm), which is partly attributed to the better light
capture ability of the first two USQs, such as their higher ε and wider absorption spectrum. Moreover, the
difference of Jsc may also be related to the carrier mobility and film morphology of the active layer.
Therefore, the hole mobility of the pure films and the blend films were tested and calculated by the space
charge limited current (SCLC) model with the device structure of ITO/MoO3 (8 nm)/USQ or USQ:PC71BM
(1:3)/MoO3 (8 nm)/Au (100 nm). The corresponding J-V curves are shown in Fig. 5. For pure films, the hole
mobility of BIDPSQ and BIDPOMeSQ are 1.54 × 10-6 cm2 V-1 s-1, 2.50 × 10-6 cm2 V-1 s-1, respectively,
which is much lower than those of IDPSQ (1.22 × 10-5 cm2 V-1 s-1) and IDPOMeSQ (2.33 × 10-5 cm2 V-1 s-1),
which may be mainly due to the much closer intermolecular pack of the latter two USQs, consistent with the
conclusion of single crystal analysis. Moreover, the methoxy-containing systems display larger hole mobility
than the corresponding methoxy-free systems, which is also consistent with the previous single crystal
analysis. For the blend films, the hole mobility of BIDPSQ and BIDPOMeSQ is 3.14 × 10-6 cm2 V-1 s-1, 7.20
× 10-6 cm2 V-1 s-1, respectively, which is also lower than those of IDPSQ and IDPOMeSQ (4.14 × 10-6 cm2
V-1 s-1, 1.12 × 10-5 cm2 V-1 s-1). Since the high hole mobility facilitates the charge transfer, the Jsc can be
effectively increased. Therefore, this is one of the reasons why the IDPSQ and IDPOMeSQ-based devices
show much higher Jsc than the BIDPSQ and BIDPOMeSQ-based devices.

a)
1
0.01
1E-4
1E-6
b)
1
0.01
1E-4
1E-6
USQ:PC71BM=1:3
BIDPSQ
BIDPOMeSQ
IDPSQ
BIDPSQ
BIDPOMeSQ
IDPSQ
IDPOMeSQ
IDPOMeSQ
1
2
3
4
5
1
2
3
4
5
Voltage (V)
Voltage (V)
Fig. 5. Current density-voltage characteristics of hole-only single-carrier devices using USQ and
USQ:PC71BM (1:3) as active layer.
Then, atomic force microscopy (AFM) measurements were conducted on the blend films of USQ:PC71BM
(1:3). AFM analysis (Fig. 6) demonstrates that all of the blend films exhibit a smooth surface morphology
with small root-mean-square roughness (δRMS = 0.20 ~ 1.59 nm). However, IDPSQ vs. BIDPSQ,
IDPOMeSQ vs. BIDPOMeSQ, the first two have relatively smaller δRMS, which is beneficial to form good
contact between the active layer and the electrode, and also facilitates charge transfer. Furthermore,
transmission electron microscopy (TEM) measurements were carried on the blend films of USQ:PC71BM
(1:3). As shown in Fig. 7, the phase separation size of the YZ-type USQ (IDPSQ and IDPOMeSQ) systems
is significantly smaller than that of the XZ-type USQ (BIDPSQ and BIDPOMeSQ) ones, which is attributed
to the intramolecular hydrogen bond interactions in IDPSQ (IDPOMeSQ) enhancing their compatibility with
PC71BM, according to our reported research results.[45] As the exciton diffusion length of SQs is generally
quite short (1 ~ 10 nm),[35] the relatively smaller phase separation size of IDPSQ and IDPOMeSQ blend
systems facilitates the diffusion of excitons to the D/A interface for dissociation, thereby improving Jsc.
Therefore, this is also an important reason that the EQE values of the IDPSQ- and IDPOMeSQ-based
devices in the absorption region of the PC71BM (300 ~ 500 nm) are still much higher than that of the
BIDPSQ- and BIDPOMeSQ-based devices. Finally, the IDPSQ- and IDPOMeSQ-based devices showed
much higher Jsc than that of the BIDPSQ- and BIDPOMeSQ-based ones.

Fig. 6. The AFM height map of BIDPSQ, BIDPOMeSQ, IDPSQ, and IDPOMeSQ blend films with
USQ:PC71BM = 1:3.
Fig. 7. The TEM images of BIDPSQ, BIDPOMeSQ, IDPSQ, and IDPOMeSQ blend films with
USQ:PC71BM = 1:3.
In addition, the PCE of BIDPOMeSQ-based device is slightly lower than that of BIDPSQ-based device,
but the PCE of IDPOMeSQ-based device is slightly higher than that of IDPSQ-based device. It can be seen
that the introduction of methoxy groups has different effects on the photovoltaic performance of different
skeletal molecules, but the overall change is not large.
Upon thermal annealing at 90 °C for 10 min, the photovoltaic performances of the four devices were
improved, with PCE values of 1.00% for the BIDPSQ-based device, 0.84% for the BIDPOMeSQ-based
device, 3.30% for the IDPSQ-based device, and 3.93% for the IDPOMeSQ-based device. It is worth
mentioning that even though IDPSQ and IDPOMeSQ show wide band gap, due to their deep HOMO level,
tight intermolecular packing and small phase separation size, the corresponding devices still achieve PCEs
of > 3.20%. Especially, the IDPOMeSQ-based device even obtains a highest PCE of 4.06%, which is
comparable to or even higher than the PCEs of some reported SQ-based devices with low band gaps.[28, 31,

43, 45, 47, 48] The photoelectric parameters of some squaraine donor materials with low band gap and their
corresponding photovoltaic performances are summarized in Table S2. By comparing these data, it can be
found that the greatest advantage of IDPSQ and IDPOMeSQ is their much lower HOMO energy level (<
-5.30 eV), which is significantly beneficial to obtain high Voc, thereby promoting higher PCE. It shows that
the YZ-typed molecular skeleton of IDPSQ and IDPOMeSQ is beneficial to construct highly efficient
unsymmetrical squaraine donor materials with low HOMO energy levels. In addition, if another acceptor
material with a low band gap matched to the IDPSQ and IDPOMeSQ is selected to prepare the active layer,
the device may achieve more excellent photovoltaic performance.
4. Conclusions
Four D-A-D’ USQs with XZ-type or YZ-type molecular skeletons, containing diphenylamine group
with/without methoxy substituent were synthesized. By systematically comparing the differences between
YZ-type USQ (IDPSQ and IDPOMeSQ) and XZ-type USQ (BIDPSQ and BIDPOMeSQ) systems from
their single crystallographic data, absorption properties, HOMO/LUMO energy levels, carrier mobility, film
morphology etc., it finds YZ-type USQs (IDPSQ and IDPOMeSQ) display closer solid-state packing, much
lower HOMO energy level and higher hole mobility. Additionally, the intramolecular hydrogen bond
interactions in IDPSQ and IDPOMeSQ are benefit to enhance their compatibility with PC71BM, leading to
smaller phase separation domain size. Consequently, YZ-type USQs exhibit the most excellent performance
with PCE of ~ 4%, which is approximately 300% higher than those of XZ-type USQs. Conversely, the
introduction of methoxy group in USQs has little different effects on the overall photovoltaic performance.
It is worth mentioning that even though IDPSQ and IDPOMeSQ show the wide band gap, the corresponding
devices still achieve a highest PCE of ~ 4%, which is comparable to or even higher than the PCEs of some
reported SQ-based devices with low band gaps. Consequently, the YZ-type molecular skeleton and the
electron-donating diphenylamine group are very promising to construct highly efficient squaraine donor
materials.

Acknowledgements
We acknowledge the financial support for this work from the National Natural Science Foundation of China
(project No. 51573108 and 21672156) and School Foundation of Southwestern Medical University (project
No. 2018-ZRZD-002). We are grateful to the Comprehensive Training Platform of Specialized Laboratory,
College of Chemistry and the Analytical & Testing Center, Sichuan University for providing NMR, HR-MS
and single crystal diffraction. We are especially grateful to Dr. Meiju Xie of the Analytical & Testing Center,
Sichuan University for the great help in TEM measurements.
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://dx.doi.org/xxxxxxxxx.
References
[1] Polman A, Knight M, Garnett EC, Ehrler B, Sinke WC. Photovoltaic materials: Present efficiencies and future
challenges. Science 2016; 352:aad4424.
[2] Wadsworth A, Moser M, Marks A, Little MS, Gasparini N, Brabec CJ, et al. Critical review of the molecular design
progress in non-fullerene electron acceptors towards commercially viable organic solar cells. Chem Soc Rev 2019;
48:1596-625.
[3] Rasi DD, Janssen RAJ. Advances in solution-processed multijunction organic solar cells. Adv Mater 2019; 31:1806499.
[4] Dey S. Recent progress in molecular design of fused ring electron acceptors for organic solar cells. Small 2019;
15:1900134.
[5] Meng L, Zhang Y, Wan X, Li C, Zhang X, Wang Y, et al. Organic and solution-processed tandem solar cells with 17.3%
efficiency. Science 2018; 361:1094.
[6] Ma X, Luo M, Gao W, Yuan J, An Q, Zhang M, et al. Achieving 14.11% efficiency of ternary polymer solar cells by
simultaneously optimizing photon harvesting and exciton distribution. J Mater Chem A. 2019; 7:7843-51.
[7] An Q, Wang J, Zhang F. Ternary polymer solar cells with alloyed donor achieving 14.13% efficiency and 78.4% fill

factor. Nano Energy. 2019; 60:768-74.
[8] Cui Y, Yao H, Hong L, Zhang T, Xu Y, Xian K, et al. Achieving over 15% efficiency in organic photovoltaic cells via
copolymer design. Adv Mater 2019; 31:1808356.
[9] Yuan J, Zhang Y, Zhou L, Zhang G, Yip HL, Lau TK, et al. Single-junction organic solar cell with over 15% efficiency
using fused-ring acceptor with electron-deficient core. Joule 2019; 3:1140-51.
[10] An Q, Ma X, Gao J, Zhang F. Solvent additive-free ternary polymer solar cells with 16.27% efficiency. Science
Bulletin. 2019; 64 :504-6.
[11] Fu H, Wang Z, Sun Y. Polymer donors for high-performance non-fullerene organic solar cells. Angew Chem Int Ed
2019; 58:4442-53.
[12] Zhang J, Zhu L, Wei Z. Toward over 15% power conversion efficiency for organic solar cells: Current status and
perspectives. Small Methods 2017; 1:1700258.
[13] Collins SD, Ran NA, Heiber MC, Nguyen TQ. Small is powerful: Recent progress in solution-processed small
molecule solar cells. Adv Energy Mater 2017; 7:1602242.
[14] Cui Y, Yao H, Yang C, Zhang S, Ho J. Organic solar cells with an efficiency approaching 15%. Acta Polym Sin 2018;
2:1-7.
[15] Xu X, Yu T, Bi Z, Ma W, Li Y, Peng Q. Realizing over 13% efficiency in green-solvent-processed nonfullerene
organic solar cells enabled by 1,3,4-thiadiazole-based wide-bandgap copolymers. Adv Mater 2018; 30:1703973.
[16] Sreejith S, Carol P, Chithra P, Ajayaghosh A. Squaraine dyes: a mine of molecular materials. J Mater Chem 2008;
18:264-74.
[17] Wei G, Wang S, Sun K, Thompson ME, Forrest SR. Solvent-annealed crystalline squaraine: PC70BM (1:6) solar cells.
Adv Energy Mater 2011; 1:184-7.
[18] Wang S, Hall L, Diev VV, Haiges R, Wei G, Xiao X, et al. N,N-diarylanilinosquaraines and their application to
organic photovoltaics. Chem Mater 2011; 23:4789-98.
[19] Chen G, Sasabe, H., Igarashi, T., Hong Z., Kido, J. Squaraine dyes for organic photovoltaic cells. J Mater Chem A

2015; 3:14517-34.
[20] Ajayaghosh A. Chemistry of squaraine-derived materials: Near-IR dyes, low band gap systems, and cation sensors.
Acc Chem Res 2005; 38:449-59.
[21] Goh T, Huang JS, Yager KG, Sfeir MY, Nam CY, Tong X, et al. Quaternary organic solar cells enhanced by
cocrystalline squaraines with power conversion efficiencies >10%. Adv Energy Mater 2016; 6: 1600660.
[22] Wu J, Si C, Chen Y, Yang L, Hu B, Chen G, et al. Photovoltaic devices prepared through a trihydroxy substitution
strategy on an unsymmetrical squaraine dye. Chem Eur J 2018; 24:3234-40.
[23] Yang L, Yang D, Chen Y, Wu J, Lu Z, Sasabe H, et al. Effects of different types of unsymmetrical squaraines on the
material properties and Coulomb interactions in organic photovoltaic devices. Mater Chem Front 2018; 2:2116-23.
[24] Mayerhoffer U, Deing K, Gruss K, Braunschweig H, Meerholz K, Wurthner F. Outstanding short-circuit currents in
BHJ solar cells based on NIR-absorbing acceptor-substituted squaraines. Angew Chem Int Ed 2009; 48:8776-9.
[25] Karpenko IA, Klymchenko AS, Gioria S, Kreder R, Shulov I, Villa P, et al. Squaraine as a bright, stable and
environment-sensitive far-red label for receptor-specific cellular imaging. Chem Commun 2015; 51:2960-3.
[26] Beverina L, Ruffo R, Patriarca G, Angelis FD, Roberto D, Righetto S, et al. Second harmonic generation in
nonsymmetrical squaraines: tuning of the directional charge transfer character in highly delocalized dyes. J Mater Chem
2009; 19:8190-7.
[27] Rao BA, Yesudas K, Kumar GS, Bhanuprakash K, Rao VJ, Sharma GD, et al. Application of solution processable
squaraine dyes as electron donors for organic bulk-heterojunction solar cells. Photochem Photobiol Sci 2013; 12:1688-99.
[28] Bagnis D, Beverina L, Huang H, Silvestri F, Yao Y, Yan H, et al. Marked alkyl- vs alkenyl-substitutent effects on
squaraine dye solid-state structure, carrier mobility, and bulk-heterojunction solar cell efficiency. J Am Chem Soc 2010;
132:4074-5.
[29] Sasabe H, Igrashi T, Sasaki Y, Chen G, Hong Z, J. Kido. Soluble squaraine derivatives for 4.9% efficient organic
photovoltaic cells. RSC Adv 2014; 4:42804-7.
[30] Yang L, Yang D, Chen Y, Luo Q, Zhang M, Huang Y, et al. Unsymmetrical squaraines with new linkage manner for

high-performance solution-processed small-molecule organic photovoltaic cells. RSC Adv 2016; 6:1877-84.
[31] Chen Y, Zhu Y, Yang D, Luo Q, Yang L, Huang Y. Asymmetrical squaraines for high performance small-molecule
organic solar cells with a short circuit current of over 12 mA cm-2. Chem Commun 2015; 51:6133-6.
[32] Ohsedo Y, Saruhashi K, Watanabe H. A new family of light-emissive symmetric squarylium dyes in the solid state.
Dyes Pigments 2015; 122:134-8.
[33] Xiao X, Cheng XF, Hou X, He JH, Xu QF, Li H, et al. Ion-in-conjugation: Squaraine as an ultrasensitive ammonia
sensor material. Small 2016; 13:1602190.
[34] Gude C, Rettig W. Radiative and nonradiative excited state relaxation channels in squaric acid derivatives bearing
differently sized donor substituents: A comparison of experiment and theory. J Phys Chem A 2000; 104:8050-7.
[35] Wei G, Xiao X, Wang S, Sun K, Bergemann KJ, Thompson ME, et al. Functionalized squaraine donors for
nanocrystalline organic photovoltaics. ACS Nano 2012; 6:972-8.
[36] Santos PF, Reis LV, Almeida P, Lynch DE. Crystal structures of a benzoselenazole-derived squarylium cyanine dye
and three derivatives substituted at the central squaric ring. CrystEngComm 2011; 13:1333-8.
[37] Lin CC, Velusamy M, Chou HH, Lin JT, Chou PT. Synthesis and characterization of naphthalene diimide (NDI)-based
near infrared chromophores with two-photon absorbing properties. Tetrahedron 2010; 66:8629-34.
[38] Ning Z, Tian H. Triarylamine: a promising core unit for efficient photovoltaic materials. Chem Commun 2009;
5483-95.
[39] Yang D, Sano T, Sasabe H, Yang L, Ohisa S, Chen Y, et al. Colorful squaraines dyes for efficient solution-processed
all small-molecule semitransparent organic solar cells. ACS Appl Mater Interfaces 2018; 10:26465-72.
[40] Chen Y, Yang L, Wu J, Wang G, Huang W, Melkonyan FS, et al. Performance, morphology, and charge recombination
correlations in ternary squaraine solar cells. Chem Mater 2018; 30:6810-20.
[41] Ohsedo Y, Miyamoto M, Tanaka A, Watanabe H. Synthesis and electrochemical properties of symmetric squarylium
dyes containing diarylamine. Dyes Pigments 2014; 101:261-9.
[42] Yang D, Yang Q, Yang L, Luo Q, Chen Y, Zhu Y, et al. A low bandgap asymmetrical squaraine for high-performance

solution-processed small molecule organic solar cells. Chem Commun 2014; 50:9346-8.
[43] Yang D, Yang Q, Yang L, Luo Q, Huang Y, Lu Z, et al. Novel high performance asymmetrical squaraines for small
molecule organic solar cells with a high open circuit voltage of 1.12 V. Chem Commun 2013; 49:10465-7.
[44] Steiner T. The hydrogen bond in the solid state. Angew Chem Inter Ed 2002; 41:48-76.
[45] Yang L, Zhu Y, Jiao Y, Yang D, Chen Y, Wu J, et al. The influence of intramolecular noncovalent interactions in
unsymmetrical squaraines on material properties, film morphology and photovoltaic performance. Dyes Pigments 2017;
145:222-32.
[46] Kan B, Feng H, Wan X, Liu F, Ke X, Wang Y, et al. A small molecule acceptor based on the heptacyclic
benzodi(cyclopentadithiophene) unit for high efficient non-fullerene organic solar cells. J Am Chem Soc 2017;
139:4929-34.
[47] Chen G, Sasabe H, Sasaki Y, Katagiri H, Wang XF, Sano T, et al. A series of squaraine dyes: Effects of side chain and
the number of hydroxyl groups on material properties and photovoltaic performance. Chem Mater 2014; 26:1356-64.
[48] Yang L, Yang Q, Yang D, Luo Q, Zhu Y, Huang Y, et al. Marked effects of indolyl vs. indolinyl substituent on
solid-state structure, carrier mobility and photovoltaic efficiency of asymmetrical squaraine dyes. J Mater Chem A 2014;
2:18313-21.